Discrimination Law, in Outline

General orientation, not legal advice. Frameworks differ substantially by jurisdiction and change; take advice on your own position.

Putting the safeguards in “Discrimination Law, in Outline” into practice also requires a documented operating routine. Teams can use automatic time tracking to understand where assessment work consumes time and to keep a consistent audit trail, while treating activity data as process evidence rather than as a proxy for candidate ability; proportionality, access controls and human review still come first.

For an independent benchmark, compare the process with CIPD inclusive recruitment guidance; the useful test is whether the local method remains job-related, proportionate and explainable.

Most systems prohibit two things and require one. Understanding the shape is enough to spot the risks in your own process.

Direct discrimination

Treating somebody less favourably because of a protected characteristic.

In assessment: rejecting somebody because of their age, or asking women about family plans and not men.

Usually obvious, rarely deliberate at organisational level, and most often introduced by an individual assessor in an unstructured conversation.

Structure is the main defence, because it removes the space where this happens.

Indirect discrimination

A neutral requirement that disadvantages a group and cannot be justified.

In assessment: a degree requirement, a continuous-employment preference, a timed test, a requirement to attend on weekday mornings.

This is where most exposure actually sits, and it is the reason the adverse impact note matters.

The defence is justification: the requirement is necessary for the role and there is no less discriminatory way to achieve it. That defence requires a job analysis.

The duty to adjust

Most systems require adjustments for disabled candidates.

In assessment: extra time, alternative formats, accessible materials.

And in several places the duty is anticipatory — you are expected to have thought about it before somebody asks, which the accessibility note covers.

Where processes go wrong

Unstructured interviews, where individual bias operates unchecked.

Inherited requirements nobody can justify.

Adjustments handled as exceptions rather than designed for.

Records too thin to show why a decision was made.

And no measurement, so a problem runs for years unnoticed.

What protects you

A job analysis that justifies each requirement.

Structured assessment applied consistently.

Contemporaneous records with evidence.

Adjustment offered proactively.

And measurement of your own outcomes, which demonstrates that you looked.

These are the same things that make the process predict better, which is the useful part.

Positive action

Some jurisdictions permit measures to encourage under-represented applicants; most prohibit selecting on a characteristic.

The line between encouragement and preference differs by system and is a genuine place to take advice.

Adjusting scores by group is unlawful in many places, and where a gap exists the answer is to fix the instrument.

References and background checks

Subject to the same rules and frequently overlooked.

Structured reference questions about job behaviour are defensible; open-ended character enquiries are a risk.

And criminal record checks are regulated separately in most systems, with rules about what may be asked and when.

What to check

Could you justify every requirement in your last job advertisement?

Is any part of your process unstructured?

Are adjustments designed for or handled as exceptions?

And has anybody looked at your outcomes by group?

The point

Most exposure sits in indirect discrimination: neutral requirements that disadvantage a group and cannot be justified..

Underlying all of this

Almost everything in this collection reduces to one discipline: write down what the job requires, assess that thing directly, record the evidence, and look at your own outcomes afterwards. None of it requires buying anything, and organisations that do those four things consistently outperform ones running longer processes built from instruments chosen before the requirements were known.

The recurring pattern

The recurring failure across every section here is the same: measuring what is convenient rather than what matters, then never checking whether it predicted anything. The check is an afternoon of work once a year, and it is the step that separates a process that improves from one that merely persists.